In addition, they are more inhibited than truth-tellers: they move their feet and legs less often. Do liars give themselves away by saying too much? In some of the studies we reviewed, liars and truth-tellers spoke very briefly — sometimes all they said was just one word. In other studies, liars and truth-tellers talked 30 Bella M. Therefore, we could look at the question of whether liars are more likely to give away their lies the longer they talk. The way we addressed this question was to look at whether cues to deception were any stronger when liars and truth-tellers talked for a longer amount of time.
We had enough data to test this hypothesis for three of the cues: pitch, response latency, and response length. For all Cues to deception and indirect lie detection 31 three, we found that the more time the liars had to talk, the more obvious these cues became. Among people who talked longer, those who were lying had a significantly higher pitch than those who were telling the truth. There was also a bigger difference in how long it took liars to begin to respond to a question, compared to how long it took truth-tellers to do so, when the answers they were about to give were longer.
Finally, in the studies in which participants generally talked longer, liars were markedly less talkative than truth-tellers. Are lies more convincing when prepared in advance? Sometimes suspects know in advance that they are going to be asked certain questions. That gives them a chance to prepare their answers. Do liars sound more like truth-tellers when they can plan their answers in advance than when they cannot? We had enough data to answer this question for eight cues, and we found the clearest results for response latency.
In the earlier analyses, in which we made no distinction between people who did or did not have an opportunity to prepare their responses, we found little difference in how long it took liars to start to answer a question, compared to how long it took truth-tellers. But when we looked separately at studies in which people had no chance to plan their answers in advance, we found that liars took longer to start answering than truth-tellers did.
However, when they did have time to plan, the liars actually started their answers more quickly than did truth-tellers. Direct and indirect deception detection The review of cues to deception showed that behavioural differences between liars and truth-tellers are generally few in number and unimpressive in size. Does this mean that people are not very good at detecting deception i. Generally, accuracy at detecting deception is little better than chance. Bond and DePaulo have summarised the results of samples in which participants were shown equal numbers of truths and lies, and asked to indicate whether they thought each message was a truth or a lie.
Overall accuracy was only about 53 per cent. The vast majority of studies of skill at detecting deception use measures such as the labelling of a message as either a truth or a lie, or rating scales assessing perceived degrees of deceptiveness. These are direct or explicit measures of deception detection and they suggest that people are not very good at discriminating lies from truths. For many years, we have included in our studies of deception detection other measures in addition to the direct measures of deception 32 Bella M.
DePaulo, A. Jordan, A Irvine, and P. Laser, , Child Development, 53, p. Copyright by Blackwell Publishers. In one of the first such studies, participants talked about other people they liked and disliked, either honestly or deceptively DePaulo and Rosenthal, The perceivers who rated the videotapes of those person descriptions were asked to report not just their impressions of deceptiveness, but also their impressions of ambivalence. The perceivers rated the lies as more deceptive than the truths. So, they were accurate in the usual way. The lies seemed especially more ambivalent than did the truths DePaulo, Rosenthal, Green, and Rosenkrantz, In a subsequent study DePaulo, Jordan, Irvine, and Laser, , the same videotape of person descriptions was shown to perceivers ranging in age from sixth-graders to college students.
They were asked to make the same judgements. The results are shown in Table 2. The scores in the first column show the degree to which the perceivers rated the lies as more deceptive than the truths — a standard, direct measure of accuracy at detecting deceit. Those accuracy scores increased with age. The second column shows the degree to which the perceivers rated the speakers who were lying as seeming to have more mixed feelings than the speakers who were telling the truth.
Those numbers also generally increased with age.
The Detection of Deception in Forensic Contexts: Practitioners' beliefs about deception
With this indirect measure, accuracy occurs at an earlier developmental point. In tenth grade, when perceivers still cannot discriminate truths from lies with their direct ratings of deceptiveness, they can distinguish truths from lies with their ratings of ambivalence. Cues to deception and indirect lie detection 33 A similar finding was reported without much comment a few years later.
Hurd and Noller asked participants to talk aloud as they tried to figure out whether a message they had just heard was a truth or a lie. However, something very interesting emerged from the verbal protocols.
When participants had just heard a lie, they were more likely to entertain the possibility that the message was a lie than when they had just heard a truth. So they had the right answer — they even said it out loud! In a longitudinal study of deception detection between same-sex friends, Anderson and his colleagues Anderson, DePaulo, Ansfield, Tickle, and Green, added several similar findings to the growing set. In that study, one of the friends in each pair told true and fabricated life stories, and the other friend tried to figure out which of the stories were true and which were made up.
The results of the dichotomous measure showed that accuracy was about the same as is typically found in studies in which judgements are made by strangers. The perceivers in the study of friends were also asked to write down whatever it was that made them think that the story was truthful or deceptive. The cues that were mentioned in these open-ended responses were sorted into a number of categories.
The two that proved most important were verbal cues e. When the friends mentioned verbal cues, they were more likely to have just heard a truthful story than a deceptive one. When the cues they mentioned were visual or demeanour cues, they were more likely to have just heard a fabricated story than a true one. So the cues the perceivers wrote down to justify their explicit guesses as to whether the stories were truthful or not did significantly distinguish the truths from the lies. In addition to making explicit judgements of whether the stories were truths or lies, and describing the cues they used to decide, the perceivers answered a series of other questions about how they felt while listening to each of the stories.
For example, they indicated how comfortable they felt, how suspicious they were, the extent to which their friend the storyteller seemed to sense their suspicion, and so forth. Videotapes of the interactions between the pairs of friends were also shown to new perceivers who answered many of the same questions Anderson, DePaulo, and Ansfield, Morris The results are shown in Table 2. At the top of the table are the results of the direct measure of deception detection accuracy for the two groups of judges.
As usual, both scores were just a shade above chance. Beneath, the table shows the results of many other indirect measures. On those measures, the perceivers did respond differently when they were reporting their impressions of deceptive interactions, relative to truthful ones. The original perceivers, the friends, felt more comfortable while listening to the true stories than the lies and thought their friend who was telling the stories seemed more comfortable. They also felt more confident when listening to the true stories, and they felt that they had got enough information, which was less likely when they had heard a lie.
The original perceivers, sitting face to face with a friend who was lying to them, felt more suspicious, tried to hide their suspicions more, and thought their friend sensed their suspiciousness more, compared to when their friend was telling the truth. The raters of the videotapes also made many of the same discriminations. They thought the storyteller seemed more comfortable when telling the truth than when lying, they felt more confident themselves when they had just heard a truthful story compared to when they had just heard a lie, and they felt that they had obtained enough information from the truthful stories.
They also felt more suspicious when hearing the lies, and they thought that the original judges interacting with the sender friends also seemed more suspicious when the story was a lie than when it was true. In all of the studies of explicit and implicit deception detection conducted in our lab, the same people made both the explicit and the implicit judgements, and the participants were college students.
Vrij and his colleagues Vrij, Edward, and Bull, extended the study of indirect deception detection in two important ways: they recruited police officers to make direct and indirect judgements, and they randomly assigned the officers to make either direct or indirect judgements. The indirect judgements in this study were impressions of whether the people on the tape needed to think hard. The police officers showed at least a hint of accuracy on both measures.
They rated the liars, compared to the truth-tellers, as somewhat more likely to be lying. They also rated them as significantly more likely to be thinking hard. Although the difference between the implicit and the explicit measure was not statistically significant because of the fairly small sample size of thirty-nine officers , the effect size for implicit deception detection was more than twice as large as the effect size for the explicit deception detection.
Cues to deception and indirect lie detection 35 Table 2. Indirect deception detection Type of story Indirect cues True Lie Indirect accuracy Own comfort Perceptions of sender comfort Confidence Got enough information Own suspiciousness Thinks sender senses suspicion Tried to hide suspicion Original judges friends 7. The first was a review of confidence at detecting deceit.
DePaulo and her colleagues DePaulo, Charlton, Cooper, Lindsay, and Muhlenbruck, found that perceivers reliably report that they are more confident when the message they just rated was truthful than when it was deceptive. In the meta-analysis of cues to deception described in detail earlier in this chapter DePaulo et al. Morris objectively. For example, to assess verbal immediacy, specific linguistic features such as the use of an active vs.
To measure eye contact, the number of seconds that the liar or truth-teller looked at the listener could be recorded. However, these cues and others could also be measured subjectively, and were in some studies. For example, perceivers could be asked to record their subjective impressions of how immediate personal, direct the liars and truthtellers seemed, or to what extent the communicators seemed to make direct eye contact.
These subjective impressions are indirect measures of deception detection. When objective and subjective measures of cues to deception differed significantly in strength, it was always the subjective measures that were stronger. Subjective impressions of immediacy separated truths from lies more powerfully than did objective linguistic measures. Subjective impressions of eye contact showed liars to be more gaze aversive than did objective measures. Furthermore, subjective impressions of facial pleasantness showed liars to be more negative than truth-tellers, even though objective measures of facial pleasantness did not.
We have one last study of indirect deception detection to report. Only that one member of the couple could see the slides. The person who could see the slides lied half the time and told the truth half the time about his or her actual feelings about the attractiveness of the people in the slides. The other member of the couple tried to tell when the first member was lying or telling the truth.
On this direct measure of lie detection accuracy, partners were correct 52 per cent of the time. However, Anderson also included complete strangers in the design. The strangers were accurate at detecting the exact same truths and lies 58 per cent of the time. So the strangers were better than the romantic partners at knowing whether the person really did find the various people in the slides to be attractive.
See Table 2. The strangers were also more accurate than the romantic partners when their direct ratings of deceptiveness were made on rating scales rather than on dichotomous measures. Anderson also included a series of indirect measures. He asked the judges to indicate, for example, how confident they felt about each of their judgements of deceptiveness, whether they felt that they had got enough information, and how suspicious they felt.
On almost all of these indirect measures, both the strangers and the romantic partners were Cues to deception and indirect lie detection 37 Table 2. Positive values indicate accurate discriminations. Ratings were made on 9-point scales. Anderson, , unpublished dissertation. They could both distinguish the truths from the lies with these indirect ratings.
For example, they felt more confident when they had just heard a truth than when they had just heard a lie. They also felt that they had obtained more information when they had heard a truth, and they felt more suspicious when they had just heard a lie. Interestingly, on all of these indirect measures, the degree to which the perceivers could separate the truths from the lies was greater for the romantic partners than for the strangers! So, even though the partners did worse than the strangers on the most direct and explicit measure, which involved calling their partners liars some of the time, they did reliably better than the strangers on the more indirect measures.
In some ways, the partners were picking up on some important behaviours that the strangers were missing. Do they have any clue at all that there could be a link between these kinds of feelings and whether or not their partner is lying? And if they were clued in on this 38 Bella M. Morris clue, would it even matter? Could they use that information effectively, or would their attempts to use it undermine the process whereby they form these meaningful impressions and intuitions? And finally, the more sinister question: If they could use this information to find out who their partners really did find attractive, would they really want to know this?
Maybe they should just let sleeping frauds lie. Conclusions As we noted at the beginning of this chapter, the existing research on deception detection should instil little faith that people are particularly good at this elusive skill. Although there are some ways in which liars behave differently from truth-tellers, there are no perfectly reliable cues to deception.
Moreover, cues to deception differ according to factors such as the type of lie and the motivation for getting away with it. Practitioners working in a forensic context or any other context may be best served by using the results of our review as a source of suggestions, each of which needs to be validated within that context. For example, if detectives had a set of videotapes of actual interrogations in which suspects had been confirmed to be lying or telling the truth, those tapes could be useful in several ways.
First, the tapes could be analysed to determine whether the same behaviours that generally separate liars from truth-tellers also do so in the context of the interrogations. Second, the tapes might be useful in training detectives to detect deception more successfully. Perhaps the detectives could be trained to attend preferentially to the cues that really are valid and to ignore the ones that are misleading.
Our research also raises the question of whether deception detection would be improved if detectives were trained to monitor and use indirect cues such as their own comfort level. This remains to be seen. It is not clear whether people can successfully introspect on how they are feeling once they are aware of the relationship between indirect cues and deception. We hope that future research will shed some light on the many important questions that still need to be addressed.
Cognitive and motivational processes underlying truth bias. Unpublished doctoral dissertation, University of Virginia. Anderson, D. The development of deception detection skill: A longitudinal study of samesex friends. Personality and Social Psychology Bulletin, 28, — Cues to deception and indirect lie detection 39 Anderson, D. Beliefs about cues to deception: Mindless stereotypes or untapped wisdom? Journal of Nonverbal Behavior, 23, 67— Bond, C.
Accuracy and truth bias in the detection of deception. Manuscript in preparation. False suspicion and the misperception of deceit. British Journal of Social Psychology, 26, 41—6. Cohen, J. Statistical power analysis for the behavioral sciences rev. Hillsdale, NJ: Lawrence Erlbaum. The accuracy—confidence correlation in the detection of deception. Personality and Social Psychology Review, 1, — Age changes in the detection of deception. Child Development, 53, —9. The motivational impairment effect in the communication of deception. Yuille ed. Dordrecht, the Netherlands: Kluwer Academic.
Cues to deception. Psychological Bulletin, , 74— Journal of Personality and Social Psychology, 37, — Diagnosing deceptive and mixed messages from verbal and non-verbal cues. Journal of Experimental Social Psychology, 18, — Reprinted, Smiles while lying. Journal of Personality and Social Psychology, 54, — Granhag, P. Deception detection: Examining the consistency heuristic.
Breur, M. Kommer, J. Nijboer, and J. Reintjes, eds. Antwerpen: Intersentia. Hall, M. Detecting deception in the voice: An analysis of fundamental frequency, syllabic duration, and amplitude of the human voice. Unpublished doctoral dissertation, Michigan State University. Horvath, F. Verbal and non-verbal clues to truth and deception during polygraph examinations.
Journal of Police Science and Administration, 1, — Differentiation of truthful and deceptive criminal suspects in behaviour analysis interviews. Journal of Forensic Sciences, 39, — Hurd, K. Decoding deception: A look at the process. Journal of Nonverbal Behavior, 12, — Inbau, F.
Criminal interrogation and confessions 4th edn. Gaithersburg, MD: Aspen. Chichester, England: John Wiley. Morris Vrij, A. Individual differences in hand movements during deception. Journal of Nonverbal Behavior, 21, 87— Legal and Criminological Psychology, 6, — Analysis of statements of victims, witnesses, and suspects. This is typically — although not exclusively — the case in sexual-abuse cases. The fundamental question to be answered in evaluating a witness statement is whether or not and to what extent the statement is a correct and complete description of the event in question.
Answering this question first requires the identification of potential sources of incorrect accounts. Based on this information appropriate diagnostic procedures and techniques can be applied in order to assess the probability of the correctness of the statement. An account may differ from the facts for two possible reasons see Figure 3. Furthermore, in both cases one can think of stable personal or of transient situational factors as the major cause for incorrect accounts. For example, sensory or intellectual deficiencies may prevent a witness from perceiving or reporting certain events.
On the other hand, darkness or lack of attention may result in incomplete perceptions in a certain situation. Of course, even people who are in general very honest may, under certain circumstances, be motivated to give an incorrect account and vice versa. All these approaches have in common the focus on indicators of deception. Rather than attempting to detect deception by looking for behavioural or physiological correlates of deception this approach focuses on variables which are associated with the truthfulness of an account. The underlying idea or theoretical framework applies as well SVA and the detection of the truth 43 to adults as it does to children.
Furthermore, this method is not limited to the assessment of accounts on sexual abuse. There is now a substantial body of empirical research that has evaluated the validity of SVA with children of different ages as well as with adolescents and adults and with accounts on a large variety of topics Vrij, This chapter will describe the basic ideas underlying SVA, the diagnostic procedure and the techniques to assess the truthfulness of a statement, empirical research on SVA and the limits of this technique. What is Statement Validity Analysis?
SVA is a comprehensive procedure for generating and testing hypotheses about the source of a given statement. It includes methods of collecting data which are relevant with regard to the hypotheses in question, techniques of analysing these data, and guidelines for drawing conclusions regarding the initial hypotheses. The basic question to be answered is: what is the source of this statement?
Does the statement describe personal experiences of the witness or does it have another source? SVA is not a psychometric test nor is it a kind of mechanistic checklist approach. Like many other forms of medical, psychiatric, or psychological assessment it is a combination of psychometric tests, semi-structured interviews, systematic analysis techniques, and clinical judgement.
These books may interest you
SVA comprises several components see Table 3. Case-file analysis A proper interview of the witness can hardly be done without any background information about the witness e. This information is utilised to generate hypotheses about the source of the statement, possible reasons for false accusations, etc.
Thus, the first stage of SVA is not, as it has been pointed out by some authors, a semi-structured interview, but a thorough analysis of the case file. Generation of hypotheses One of the most important and indeed crucial components of SVA is a careful generation of hypotheses, that is, assumptions about potential sources or origins of the given statement. Everything else, the whole assessment procedure, the data to be collected, and the particular evaluation strategies depend on these hypotheses.
This overall hypothesis is then broken down into more specific assumptions about potential sources of a presumed incorrect statement. SVA and the detection of the truth 45 Complete fabrication The event that is described in the statement may be completely fictitious, that is, it never happened to the witness. The witness may have intentionally invented the statement in order to achieve certain personal goals e. According to this hypothesis the complete statement has been intentionally fabricated.
Partial fabrication It is possible that large parts of the statement are correct accounts of a certain course of events, whereas some crucial elements are fabricated. This sometimes happens in cases of alleged rape. The accused person may admit contacts with the witness, even a sexual relationship and an encounter at the location mentioned by the witness.
It is, however, denied that any sexual activity happened without explicit or at least implicit consent. Obviously, under such circumstances the lying witness would not have to fabricate the complete statement. Only those sequences of the statement that describe his or her resistance and any resulting specific forms of interaction would have to be fabricated. Incorrect transference Another form of partial fabrication can occur if the witness has indeed experienced the event in question or at least something similar to it , but with another person than the one who is accused.
For example, a child may have had previous experiences of sexual abuse or a woman may have been raped before. In such a case the event itself is reported from memory while the particular relations to the accused person, the time and location would have been fabricated. Instruction by others It is possible that a child witness has been instructed by another person, usually an adult, to claim incorrectly that something has happened.
This may happen, for example, in custody conflicts after a divorce. The statement would then not be fabricated by the child witness, but by an adult instructor, and communicated to the child. It is therefore possible that the witness has experienced suggestive influences and that these have produced a statement which is now perceived as a truthful account by the witness.
Mental illness A witness may have problems distinguishing fact and fantasy as a symptom of mental illness e. He or she may then take fictitious events as real and subjectively give a truthful account. These hypotheses are just some examples, not an exhaustive list, of potential sources of incorrect statements. On the other hand, not all of these hypotheses are relevant in any specific certain case. For example, there may not be any indication for a mental illness nor opportunities for any suggestive influences. Setting up a diagnostic strategy and deciding upon assessment methods The hypotheses which have been generated based on the results of the case-file analysis are now used to set up a diagnostic strategy, to decide which assessment techniques e.
A detailed description of the location may be of minor diagnostic significance only if it is not disputed that the witness has been there. If the hypothesis of incorrect transference has been proposed, the interview would focus particularly on those elements that relate the described event to the accused person and to the time and location where it presumably happened because these are the elements which would have been fabricated according to this hypothesis.
These are, therefore, the diagnostically relevant parts of the statement. If a hypothesis of mental illness has been proposed, it may be necessary to evaluate hospital files, to ask the court to hear the report of the psychiatrist who has treated the witness, or to apply clinical assessment SVA and the detection of the truth 47 procedures before interviewing the witness. Similarly, if the case file reveals that a witness has been inappropriately interviewed before, has been in psychotherapy, etc. In this case a thorough analysis of the circumstances of the first disclosure, the responses of others to these reports, the type of conversations and questions that have been asked, the types of therapy, etc.
In order to evaluate the hypothesis of a complete or partial fabrication of the statement, a judgement has to be made as to whether or not the witness would have been able to fabricate this statement. This is done using the CBCA criteria. These examples demonstrate that the assessment procedure and techniques largely depend on the hypotheses that have been generated based on the case file analysis. It is therefore obvious that an interview of the witness cannot reasonably be the first stage of SVA.
Indeed, in light of some hypotheses it may not even be advisable to interview the witness at all. If, for example, the witness has participated in a highly suggestive therapy over a considerable period of time, it may be impossible to generate an uncontaminated statement. In such a case interviewing the witness may be useless. Examination of the witness When the diagnostic strategy has been set up and relevant assessment techniques have been selected, the witness is examined.
If it is a child witness the examination usually begins with a history-taking interview with the parents. Such an interview helps to determine whether or not critical life events have happened to the witness e. It also provides additional background information about a witness, his or her developmental stage, diseases, etc. The exact sequence of the application of further assessment procedures largely depends on the particular circumstances and may vary from case to case. However, the examination almost always includes an interview of the witness.
Sometimes psychometric tests e. In other cases other sources of information personal impression during the examination, school reports, and results of previous tests are available so that no additional testing is required. The details of this special form of investigative interview have been described elsewhere e.
This interview is investigative, not therapeutic. Indeed, therapeutic interviews should, whenever possible, be postponed until after the investigative interview s in order to avoid any suggestive contamination of the statement. It is important to stimulate the witness to give a free report of the event in question before asking any questions. The questioning part of the interview should start with open-ended questions. Of course, a witness should not be prompted to produce certain contents which relate to any CBCA criteria.
Most of the CBCA criteria depend on proper interviewing. Police interviews, even if they are available as verbatim transcripts, are usually not suitable for a CBCA because they are often highly structured by the interviewer and focused on details which are relevant for the criminal investigation, while other details are more or less neglected e.
Furthermore, police interviewers have rather different hypotheses in mind from the ones outlined above. Consequently, their interviews will focus on different aspects of an incident. Katofsky and Eggers have shown that, although police interview transcripts can be coded reliably, the proportion of correct classifications as truthful or fabricated is considerably lower than the ones reported in other studies e. As it has been pointed out above, any components of a statement are meaningful only if they are able to distinguish between truthful and partially fabricated statements.
Applying CBCA criteria without any selective differentiation can lead to completely meaningless results. This again demonstrates the importance of appropriate hypotheses before the examination begins. This hypothesis was originally stated by Undeutsch He has assumed that fabricated accounts, among other things, contain fewer details, are less vivid and less colourful than statements derived from original memory.
The SVA and the detection of the truth 49 latter can be related to impression management theory Tedeschi and Norman, The cognitive part of the hypothesis states that, given a certain level of cognitive and verbal abilities, only a person who has actually experienced an event will be able to produce a statement with the characteristics that are described in the CBCA criteria see Table 3.
It is, for example, deemed unlikely that a child who fabricates an account will deliberately invent unexpected complications that allegedly happened during the incident. The impression-management component relates to motivation and social behaviour. It is assumed that lying is a goal-directed behaviour and that a person who deliberately invents a story wants to be perceived as honest in order to achieve his or her goals. Therefore, the person is likely to avoid behaviours which, in his or her view, may be interpreted as clues to deception. For instance, if a liar believes that admitting lack of memory will undermine his or her perceived credibility, he or she will try to avoid such behaviour.
This impression-management approach assumes that people have a common stereotype about the typical behaviour accompanying a lie. Provided that a particular behaviour can be sufficiently controlled it is expected that a liar, in order to conceal his or her lie, attempts to avoid such behaviour.
Based on these hypotheses a number of content features have been described which are likely to occur in accounts of genuine experiences but less likely to be found in fabricated statements. Thus, these criteria can be used to test the hypothesis of partial fabrication of the statement. Similar lists of content characteristics have been published by Arntzen and Littmann and Szewczyk Logical structure requires that the statement contains no contradictions or logical inconsistencies.
It is important to note that neither the account of unusual details criterion 8 nor the report of unexpected complications criterion 7 interfere with the logical consistency. Furthermore, logical consistency is not the same as plausibility. A statement may sound implausible but nevertheless be logically consistent. Unstructured production. Sometimes a witness reports various elements of an incident in an unsystematic, chronologically disorganised manner. They follow their momentary associations rather than a logical and chronological sequence. This is extremely difficult without any genuine memories and, therefore, unlikely to occur in fabricated statements.
Quantity of details. The event, the location and surroundings, and the people involved are described in great detail. Logical structure 2. Unstructured production 3. Quantity of details Specific contents 4. Contextual embedding 5. Descriptions of interactions 6. Reproduction of conversation 7. Reporting of unexpected complications during the incident Peculiarities of content 8.
Unusual details 9. Superfluous details Accurately reported details misunderstood Related external associations Accounts of subjective mental states Spontaneous corrections Admitting lack of memory Self-deprecation Pardoning the perpetrator Offence-specific elements Details characteristic of the offence Contextual embedding means that the event in question is described as related to particular locations, time schedules, personal relationships to the accused and other persons before and after the incident.
Descriptions of interactions. Naturally occurring events are usually characterised by a sequence of actions and reactions. The reactions or responses can be explained psychologically as consequences of previous actions. Particularly for child witnesses, it is difficult to fabricate psychologically comprehensible sequences of interactions. The reproduction of conversation differs from the previous criterion in one important aspect. It means that the witness reproduces conversations between different persons in a kind of role-play, thereby using the particular speech behaviour, vocabulary, etc.
SVA and the detection of the truth 51 Reporting of unexpected complications. Sometimes the course of events is interrupted by unexpected complications and obstacles. The range of these complications can extend from an unforeseen interruption or difficulty to the spontaneous stopping of the event before its logical completion.
Unusual details refer to particular elements or details of a statement which are unexpected and surprising, e. Superfluous details. Details are superfluous if their reporting is not strictly necessary for the description of the incident in question. Accurately reported details misunderstood. It means that details and actions are reported which are obviously not understood by the child in their particular meaning and function as sometimes indicated by erroneous interpretations of correctly described observations.
Related external associations. It means that, for example, a witness describes conversations with the perpetrator that refer to earlier events which are in some way related to the incriminated actions. For example, in a report about an incestuous relationship, such associations would be reports of the witness daughter about conversations with the perpetrator father in which they discussed the sexual experiences of the daughter with other partners. Accounts of subjective mental states. The accounts of subjective mental states include the description of feelings such as disgust or fear and the reporting of crime-related cognitions like, for example, thinking about how to escape.
However, the mere mentioning of a particular emotion like, for instance, fear would not be sufficient to fulfil this criterion. Spontaneous corrections. A witness corrects or modifies previous descriptions without having been prompted by the interviewer. Admitting lack of memory. A witness expresses concern that he or she may not remember all relevant details, that the description of particular details may be incorrect, etc. The witness indicates that part of his or her description sounds odd, implausible, unlikely, etc.
The witness mentions personally unfavourable, selfincriminating details. Pardoning the perpetrator. The witness excuses the accused person for his or her behaviour. Details characteristic of the offence. The witness describes elements which are typical for this type of crime but, on the other hand, are counter intuitive for the general public or discrepant to everyday knowledge or stereotypes. The first stage is the analysis of the statement with regard to the CBCA criteria.
As such, these data are completely meaningless. Remember that the question to be answered with the help of CBCA is whether or not this witness would have been able to fabricate this statement with the content qualities identified in the account without having personally experienced the event described. Obviously an adult witness with good cognitive and verbal abilities, creativity, general knowledge, etc. As a consequence, the same content-analysis results i.
The Detection of Deception in Forensic Contexts - PDF Free Download
Therefore, in a second step, the results of the content analysis i. A certain criterion is then and only then fulfilled or more or less pronounced in a statement, when this particular content feature is unlikely to occur in a statement fabricated by this witness. For example, a witness in a sexual-abuse case reports that the accused person has threatened the victim because of the possibility that he or she might tell others about the incident. If, for example, the witness knows from the literature, magazines, TV, teachers, etc. It does not discriminate between fabricated and truthful accounts and is, therefore, diagnostically meaningless.
There are, of course, no quasi-psychometric or statistical rules which can be applied in order to generate these standards or references. If CBCA leads to the conclusion that a fabrication of this statement by this witness is unlikely, then this fabrication or partial fabrication hypothesis is rejected. Note that this is only an intermediate result. As long as there are other potential sources of the statement e.
It only means that it is unlikely to have been fabricated. If, on the other hand, the CBCA results are not sufficient to reject the fabrication hypothesis, no further analyses are necessary. It will then not be possible to conclude reliably that the statement is based on own experiences. It is, of course, possible that some events which are described in a statement have been experienced whereas others are fabricated.
CBCA is then applied to each of these elements separately. This procedure may lead to a rejection of the fabrication hypothesis for some of the reported events but not for others e. Thus, there is no general assumption underlying CBCA that a statement is either completely truthful or completely fabricated, as Vrij has pointed out. Consequently, the proposition that there is no procedure to distinguish between experienced and non-experienced portions within the same account Vrij, is not correct. Analysis of consistency If the witness has been interviewed repeatedly and if these interviews are well documented in the case file, it is possible to compare the statements in order to identify any discrepancies.
Contrary to widespread assumptions it is not required that two or more accounts do not differ in any detail in order to be judged as credible. Indeed it has never been claimed that truthful accounts given at different times have to be identical e. In line with this position recent research has shown that truthful accounts given at different times are more or less inconsistent see, e. If, however, substantial discrepancies are found in the description of the core of an event, then this would at least require an explanation.
Evaluation of additional hypotheses If the partial fabrication hypothesis is rejected, it is still possible that the statement does not describe own experiences. The statement could be the result of previous suggestive interviews or psychotherapeutic interpretations. Therefore, CBCA cannot help to decide upon a suggestion hypothesis. If, for example, the first disclosure was made immediately after an alleged incident and the witness was interviewed shortly thereafter by the police, there may not have been any room for suggestive influences, and this hypothesis can thus be rejected.
In other cases many conversations, formal and informal questionings, and perhaps psychotherapeutic sessions may have preceded the first documented interview by the police. It will then be much more difficult, or perhaps even impossible, to rule out the possibility of suggestion as the major source of the statement.
Other hypotheses like, for example, the possibility of a mental disease and a related inability to distinguish fact from fantasy are usually evaluated before the witness is interviewed and a CBCA is performed. Final judgement of statement credibility The final conclusion on statement credibility is the result of evaluating all relevant hypotheses about potential sources of the statement.
Data are collected and evaluated in light of these hypotheses. A hypothesis which is incompatible with the diagnostic findings is considered to be insufficient and thus rejected. This process is continued until either all hypotheses which suggest that the statement is incorrect are rejected leaving only the hypothesis of own experiences as the source of the statement or until an alternative hypothesis i. Note that this does not automatically mean that the witness is lying. It is quite possible that the statement is a truthful account of own experiences but due to inconclusive, ambiguous data or suggestive questioning alternative hypotheses could not be rejected.
SVA is, therefore, a one-dimensional assessment procedure. If all other hypotheses are rejected, the hypothesis of own experiences as the source of the statement remains the only plausible explanation. No statement available Since the major part of SVA is the content analysis of a statement, it cannot be applied if no statement is available. This can, for example, happen with very young children or with people with learning disabilities. Insufficient material In particular circumstances a detailed statement may be available but it may nevertheless not contain enough material that is suitable for a CBCA.
This happens sometimes in cases of alleged rape where the accused person admits sexual intercourse at the location in question but denies that it happened without the consent of the witness. The basic assumption underlying CBCA and SVA is whether or not the witness would have been able to fabricate the statement without having personally experienced the event in question. Therefore, the diagnostically relevant part of the statement may, for example, be reduced to the description of resistance, violence, and related interactions.
This may then not be sufficient to apply CBCA. It is also possible that the event in question was rather short, simple, without any special elements e. If, in addition, the time interval between the event in question and the interview is rather long, the resulting statement may be too short to allow a thorough content analysis. Previous suggestive questioning Research on suggestion has shown that repeated suggestive questioning might produce statements which are very similar to accounts of own experiences for reviews, see, e.
CBCA and SVA have been constructed to discriminate truthful from fabricated statements, not to discriminate truthful from suggested accounts. Similar problems can arise from previous psychotherapeutic treatment. Some forms of psychotherapy have a considerable suggestive potential e. Coaching There is now increasing empirical evidence which shows that witnesses can be coached to produce statements with CBCA-relevant contents. Inappropriate interviewing It has been emphasised that a proper interview of the witness is required in order to apply CBCA.
It is not only important to follow the general guidelines for good practice in interviewing e. In addition, the interview has to be focused on those aspects of the event in question that are relevant for evaluating the initial hypotheses. We have found, SVA and the detection of the truth 57 for example, that although the CBCA criteria could be reliably coded in police interviews, the proportion of correct classifications was rather poor and well below the results of other studies Eggers, ; Katofsky, In the light of these results it is at least questionable whether CBCA can reliably be applied to statements which have been generated using insufficient interview techniques.
If they are not, SVA will probably not be of much help in evaluating the truthfulness of a statement. Almost all of the more recent studies have evaluated the classification accuracy of CBCA rather than the validity of the outcome of the whole SVA procedure see Vrij, , for a review. Although interrater reliability of coding of CBCA criteria is an essential prerequisite for valid classifications, systematic research on this issue is even more recent.
Interrater reliability Since CBCA is a major part of Statement Validity Analysis, a reliable coding of the criteria is essential for valid classifications of statements as fabricated or truthful. This will result in high proportions of error variance in the outcome of content analysis.
As a consequence, it will be less likely to find significant differences between truthful and fabricated statements. Vrij found in a review of thirty-seven studies on CBCA and SVA that in most of these studies the differences in CBCA scores were in the predicted direction but sometimes missed the significance levels. It is quite possible that some of the failures to significantly discriminate truthful from fabricated statements were due to insufficient training of the coders.
Reading of relevant book chapters, research papers, and detailed descriptions of CBCA criteria 2. Lectures on observation and rating methods and common rating errors 3. Presentation of extended criteria descriptions; discussion of positive and negative examples i. Discussion and evaluation of homework, rehearsal of criteria descriptions, identification of criteria in transcribed statements, homework assignment 5.
Discussion and evaluation of homework, introduction of rating scale, presentation and discussion of examples, homework assignment 6. Discussion and evaluation of homework, application of rating scale to transcribed statements, discussion of consistencies and discrepancies, comparison with expert ratings, homework assignment 7. Discussion and evaluation of homework, discussion with experts, presentation and discussion of examples of transcribed statements Table 3. The programme extends over three full weeks with five days a week and 7—8 hours per day during week 2 and 3 and comprises several components see Table 3.
The agenda of a typical training day is shown in Table 3. In one of these studies Petersen, three groups of coders read several articles and book chapters SVA and the detection of the truth 59 on CBCA, or attended a university seminar on statement credibility and SVA, or participated in the training programme. A control group received no treatment at all. Different selections of several transcribed statements were coded according to the CBCA criteria before and after training.
In addition, raters re-coded the same statements again after a delay of three weeks in order to assess the intrarater agreement or stability of the codings. A five-point scale was used rather than three-point scales which have been applied in a number of other studies because it proved to be more sensitive for smaller differences between truthful and fabricated statements. A correlation coefficient, on the other hand, underestimates the actual agreement when the variance is small e.
Training significantly increased the average interrater agreement across all criteria and all statements from around above 40 per cent in the baseline condition to almost 70 per cent whereas no improvement was achieved in the reading, seminar, and not training control group. These results show that just reading texts on CBCA and SVA or attending a seminar does not increase interrater agreement substantially compared to a no-treatment control group.
A retest shows that these effects appear to be fairly stable over a delay of three weeks with an average intrarater agreement around 75 per cent, as compared to around 50 per cent in all other groups. Six similar studies using different groups of raters and different statements lead to similar results. When we looked at the interrater reliability for each individual CBCA criterion we found that the majority of discrepancies are not larger than one point on a five-point scale.
For sixteen out of the nineteen reality criteria more than 90 per cent of the ratings did not differ more than one point on a five-point rating scale. Agreement between raters never fell below 80 per cent.
Taken together these results show that CBCA criteria can be coded at reliability levels which are comparable to the ones reported for common personality questionnaires. Without such training a high level of error variance may cover differences of small magnitude between fabricated and truthful statement. He discovered that the expected differences were found in CBCA scores between liars and truth-tellers in all experimental research paradigms actual involvement, watching a video or statements derived from memory.
The findings published so far also support the assumption that CBCA is not restricted to statements of children who have been victims of sexual abuse but can also be used with adults and with different types of crime. Furthermore, the hypothesis that truth-tellers will obtain higher total CBCA scores than liars was examined in twelve of the studies reviewed by Vrij , and in eleven out of these studies the hypothesis was supported.
The differences were always in the expected direction but they were not always statistically significant. However, this may be due to insufficient training or to insensitive three-point rating scales. The proportion of correct classifications in the studies reviewed by Vrij varied from 65 to 90 per cent with an average accuracy rate for truths of 73 per cent and for lies of 72 per cent. Recently the German Supreme Court has ruled that SVA is the method of choice whenever psychological experts submit their opinion regarding the credibility of a statement to a court.
Similar rulings exist, for example, in Switzerland. Obviously, this acceptance is the result of strongly increased empirical, mainly experimental, research during the past two decades. This does not mean that SVA claims to be the ultimate solution to all legal problems. Like any other diagnostic procedures it has its strengths and weaknesses and its limitations.
However, the question is not whether or not SVA is perfect in distinguishing truthful from incorrect accounts. Instead, the crucial question is whether or not SVA is significantly better in identifying the credibility of a statement than other judgement procedures which are also available. Polygraph techniques are not always applicable and in many countries not admissible as evidence.
Global judgements of credibility by laypersons almost invariably result in accuracy rates around chance level. Research on non-verbal indicators of deception has in the past primarily focused at group differences and has not yet sufficiently proved to be applicable in the assessment of individual statements.
Deception detection and reality monitoring: A new answer to an old question? Bender, and T. Bliesener eds. Berlin: DeGruyter. Arntzen, F. Psychologie der Zeugenaussage. Bull, R. Obtaining information from child witnesses. Memon, A. Vrij, and R. Bull eds. Maidenhead, England: McGraw-Hill. A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37— Eggers, J. Erdmann, K. Induktion von Pseudoerinnerungen bei Kindern. Regensburg: S. Roderer Verlag. Fleiss, J. Measuring nominal scale agreement among many raters. Psychological Bulletin, 76, — Description Table of Contents Product Details Click on the cover image above to read some pages of this book!
Industry Reviews 'It will be a valuable resource for forensic practitioners, academics and students.
- The Spirit of Hindu Law.
- Other Subject Areas?
- Post Comment.
- The Behavioral Neurology of White Matter.
Introduction Research on deception detection: past and present Lie-Detection Techniques Discerning lies from truths: behavioural cues to deception and the indirect pathways of intuition Statement validity analysis and the detection of the truth Reality monitoring and detection of deception The psychophysiological detection of deception Special Problems Facing a Lie-Catcher Lies travel: mendacity in a mobile world Coping with suggestion and deception in children's accounts The detection of false confessions Crime-related amnesia as a form of deception Enhancing Lie-Detection Accuracy Practitioners' beliefs about deception Training to Detect deception from behavioural cues: attempts and problems The wizards of deception detection Guidelines to catch a liar Conclusions Research on deception detection: future challenges Table of Contents provided by Publisher.
All Rights Reserved. In Stock. Unbelievable The chilling true story of the woman nobody bel Company Law Perspectives 3rd Edition. Lawyers' Professional Responsibility 6th Edition. Principles of Criminal Law. Concise Australian Commercial Law 5th Edition. A Practical Introduction to Taxation Law Australian Criminal Justice Fifth Edition.
Statutory Interpretation. Ethics and Law for Australian Nurses 3ed. Administrative Law Context and Critique. Cases on Torts. Stewart's Guide to Employment Law 6th Edition. View Wishlist. Our Awards Booktopia's Charities. Are you sure you would like to remove these items from your wishlist?